FULL STACK SOFTWARE ENGINEER

From zero to
working software.

I’m Eduardo Escamilla, a software engineer based in Monterrey. I build web and mobile products with React, Next.js, React Native, Node.js and Python, and I’m particularly comfortable with greenfield development. My background also includes mathematics, quantitative modeling and machine learning.

Eduardo Escamilla
Eduardo EscamillaMonterrey, Mexico

Experience at a glance

Explore the projects and professional experience behind each area.

01Greenfield development ↗Idea or design → production
02Production engineering ↗Web · mobile · backend · cloud
03Quantitative background ↗Mathematics · modeling · ML
04Recent AI work ↗LLMs · automation · structured outputs

SELECTED WORK

Work that shows how I approach problems.

01 / FLAGSHIP

FORMAL LOGIC · STEM

ProofGym

AI can give you an answer in seconds. ProofGym gives you a reason to work through one yourself. Practice building arguments and tackling STEM problems where the reasoning matters.

PROJECT EXPLORER

Explore ProofGym with AI

AI can supply an answer. ProofGym asks you to build the reasoning. Explore why this project exists, how its logic and STEM exercises work, or the code behind them.

View code · ProofGym · lib/validator.ts · isAccessible
export function isAccessible(
  targetId: number,
  fromId: number,
  allLines: ProofLine[],
  closedSubproofs: ClosedSubproof[],
): boolean {
  if (targetId >= fromId) return false

  const targetLine = allLines.find(l => l.id === targetId)
  if (!targetLine) return false

  // A closed subproof's interior is inaccessible to lines OUTSIDE it.
  // Lines inside the same subproof can still reference each other freely.
  for (const sp of closedSubproofs) {
    if (targetId >= sp.startId && targetId <= sp.endId) {
      // Target is inside this closed subproof.
      // Only inaccessible if fromId is OUTSIDE the same subproof.
      const fromIsInside = fromId >= sp.startId && fromId <= sp.endId
      if (!fromIsInside) return false
    }
  }
  return true
}

AI guide to this project. Questions are processed by OpenAI.

1   P → Q2   P3   Q    →E 1,2   ✓⊢
ProofGym · App walkthrough · 3× speed

THE PROBLEM

I wanted a place to practice independent reasoning in an age of instant answers: question assumptions, follow an argument, and understand why a conclusion holds.

WHAT I BUILT

A formal-logic workspace for building proofs one step at a time, alongside STEM and critical-thinking challenges. Choose an exercise, construct your reasoning, and use feedback to reconsider a step when it does not follow.

WHY IT MATTERS

AI opens up new ways to learn. ProofGym complements those tools with hands-on reasoning practice: time to explore an idea, test an argument, and reach your own conclusion.

expression→rule→references→validation→feedback
02 / PRODUCTION

REACT NATIVE · GEOLOCATION · AWS

GPS → geographic region

A production mobile feature that classified a user’s GPS position into configured geographic zones represented by KML polygons.

GPS
lat / lon
→
KML
polygons
→
point in
polygon
→
matched
zone

CONTEXT

Part of my work on a production React Native application, React administrative system and AWS-backed services.

ENGINEERING

The feature connected device geolocation with KML polygon data so the application could determine which configured zone contained a coordinate.

03 / EXPERIMENT

HYPOTHESIS · PROTOTYPE · EXPERIMENT

KickboxWM

What if a workout challenged your memory as deliberately as your body? Recall an earlier movement cue, resist reacting to the newest one, and keep moving. An experiment in combining physical effort with cognitive practice.

PROJECT EXPLORER

Explore the KickboxWM experiment

What if a workout challenged your memory as deliberately as your body? Explore the idea behind KickboxWM, the delayed-recall movement task, and what it would take to test the hypothesis.

AI guide to this project. Questions are processed by OpenAI.

movement← independent stream →
visual memory← independent stream →
N−BACKtiming · audio cues · work/rest · session state
04 / IN PROGRESS

INTERACTIVE GAMES · MEMORY · ATTENTION

CogniTrain

Seven browser-based cognitive games covering memory, attention, processing speed, mental arithmetic, and emotion recognition. Short sessions combine adaptive challenges and progress tracking.

PROJECT EXPLORER

Explore CogniTrain with AI

Explore short game sessions, adaptive challenges, and progress tracking in a browser-based app currently in development.

AI guide to this project. Questions are processed by OpenAI.

memoryDual N-Back · memory sequences
attentioncolor-word challenges
7 GAMESshort sessions · adaptive challenges · progress tracking
CogniTrain · Gameplay preview

ENGINEERING RANGE

React ecosystemReact · Next.js · React Native · responsive interfaces
Backend & dataNode.js · Express · Python · SQL · Supabase
Cloud & deploymentAWS · GCP · Azure · Vercel · Docker · Linux
ML & LLM workPython · machine learning · Gemini · Claude · n8n

EXPERIENCE

Software shipped in professional environments.

A condensed history. Download the CV for the full version.

Download CV (PDF)

Q Bridge AI

Full Stack Software Developer / Head of Solutions

Full-stack applications and dashboards with React, Next.js, Node.js, Express, Python, SQL and Supabase. LLM workflow prototypes with n8n, Gemini and Claude. Application deployment and troubleshooting across AWS, Google Cloud, Docker and Vercel.

Avant Santé

Full Stack Software Developer

Built the company website. Separately developed a backend system with APIs, database integrations and an administration portal using React, Python, Node.js and SQL; this additional system was not deployed.

The Blue People

Full Stack Software Developer

Production React Native mobile applications, React administration tools and serverless systems using AWS and Azure. Work included authentication, real-time messaging, media, geolocation and analytics.

Independent

Software Developer & STEM Tutor

Custom .NET solutions and mentoring in mathematics, physics, programming and mathematical modeling.

Interfactura Grupo GDX

Software Developer

Python/data work, web development, CRUD applications, Linux, Docker and Git.

KXK Global

Financial Modeling

Quantitative risk models and stochastic simulations in R for financial and project risk.

University of Texas–Pan American

Game Theory Research Assistant

Worked with game-theoretic models related to seller competition, auctions and mechanism design.

TEAM / HACKATHON EXPERIENCE

CodeLaunch Houston 2022

I participated at CodeLaunch Houston 2022 as part of The Blue People development team, working in a fast-paced collaborative product environment.

BACKGROUND

Mathematical thinking.
Practical software.

My background in mathematics, game theory research, and quantitative modeling informs how I approach software and machine learning. I continue building on that foundation through structured learning and projects that put the ideas into practice.

Explore my professional development

Education

UNIVERSITY STUDIES · 2007–2010

Mathematics

University of Texas–Pan American (UTPA)

Additional coursework in Economics and Physics.

UNIVERSITY STUDIES

Computer Science

Universidad Virtual del Estado de Guanajuato (UVEG)

Research & honors

2008–2009Game Theory Research

University of Texas–Pan American · seller competition and mechanism design.Seller Competition by Mechanism Design · Damian S. Damianov
Acknowledged in the published article for contributions through comments.View publication & acknowledgment Download paper (PDF)

2013–2014Quantitative Modeling

Stochastic simulation and quantitative risk modeling in R.

2003Mexican Mathematical Olympiad

Silver Medal.

2004Regional Mathematical Olympiad · Nuevo León

1st Place Overall.

Professional Development

Completed tracks, specializations & selected courses

Structured learning in software, data engineering, and machine learning, supported by completed projects and coursework.

4 completed DataCamp tracks · Coursera specializations & courses · edX coursework

HOW I WORK

Curiosity, practice, and honest work.

I use AI as part of my workflow. This portfolio is a place to share the things I love building, problems I’ve tackled, and ideas I’m still learning. The projects here show what I’ve built and how I approach problems.

A little more about me

My interests connect mathematics, formal reasoning, software, and learning. ProofGym and KickboxWM explore different sides of that curiosity: practicing rigorous thinking and experimenting with interactive training.

CONTACT

Looking for someone who can build from scratch?

eduardo.escamillasal@gmail.com

Send me a message here. You may be asked to complete a quick spam check before confirmation.